The primary goal of this research is to understand the cause of the abnormal phenotypes in mouse embryos developing in the absence of the Dnmt1o form of the Dnmt1 cytosine methyltransferase. The central premise to be tested is that the majority of Dnmt1o-deficient embryos die because they are epigenetic mosaics of cells that have defective imprints. Consequently, they do not develop normally. This model system will allow a detailed examination of the effect of epigenetic variation on phenotype and pregnancy outcome. Aim 1: Determine the nature of the epigenetic mosaicism in embryos derived from Dnmt1o-deficient oocytes. An embryo derived from a Dnmt1(delta1o) homozygous female develops in the absence of maternally supplied Dnmt1o and exhibits a loss of DNA methylation from 50% of the normally methylated alleles of imprinted genes. Each embryo is a mosaic, and should be comprised of many different epigenetic cell types, ranging from cells with mostly normal methylation on imprinted genes to cells with nearly no methylation on imprinted genes. This prediction will be tested by measuring imprinted methylation patterns in clones of embryonic stem (ES) and embryonic fibroblast (EF) cells derived from Dnmt1o-deficient embryos. Aim 2: Establish the source of the epigenetic mosaicism in embryos derived from Dnmt1o-deficient oocytes. The loss of Dnmt1o-dependent maintenance methylation probably occurs at the fourth embryonic S phase. This hypothesis will be tested by measuring imprinted methylation patterns in pools of Dnmt1o-deficient embryos at each preimplantation stage, and in cloned embryos derived using donor nuclei from each Dnmt1o-deficient preimplantation stage. Aim 3: Determine the cause of the variable, abnormal phenotypes of Dnmt1o-deficient embryos. Dnmt1o-deficient embryos die, most likely because their imprint-defective cells are unable to contribute to normal development. This hypothesis will be tested by first defining the morphological abnormalities present in Dnmt1o-deficient embryos and extraembryonic tissues. Afterwards, clones of Dnmt1o-deficient ES cells, each a self-renewing population with a unique pattern of normal and defective imprints, will be tested for their ability to contribute to embryonic development in ES-blastocyst and ES-tetraploid chimeras.